Rhoda AI, a Palo Alto–based robotics company, has raised $450 million in Series A funding and introduced a software platform designed to help industrial robots operate in environments that change frequently, such as manufacturing and logistics facilities.
The company develops artificial intelligence software that allows robots to observe their surroundings, predict how objects and environments will change, and adjust their actions in real time. The technology is intended to enable robots to perform physical tasks in settings where layouts, materials, and workflows vary, conditions that have historically limited the use of automation.
Industrial robots are typically programmed to follow fixed trajectories and operate in highly structured environments. While newer artificial intelligence approaches allow robots to learn from data, many systems struggle when faced with unfamiliar objects, shifting workspaces, or unexpected changes during operation.
Rhoda AI’s system is designed to address this problem by training models on large volumes of video. The company said its models are initially trained on hundreds of millions of internet videos to learn patterns of motion, physical interaction, and environmental change. The models are then refined using smaller datasets collected from robotic systems so they can translate those predictions into physical actions.
The company’s software continuously observes the environment through cameras, predicts what is likely to happen next, converts those predictions into robot movements, and updates its behavior as conditions change. This process runs repeatedly in short cycles, allowing the robot to adjust to new information while it is performing a task.
Rhoda AI refers to this architecture as a Direct Video Action model, which links visual perception directly to robotic control. According to the company, the approach allows robots to learn new tasks with relatively small amounts of robot-specific training data.
The software platform, called FutureVision, serves as the company’s core intelligence system. Rhoda AI plans to use it to power its own robotic deployments and may license the technology to partners developing different robotic hardware and software systems.
The company said its technology has already been tested in production environments where robots must handle changing materials and workflows. In one manufacturing evaluation, a robotic system using Rhoda’s software completed a component-processing task in under two minutes per cycle without human intervention.
The funding will support further research and engineering development, expansion of industrial deployments, and additional customer pilot programs. Investors in the round include Capricorn Investment Group, Khosla Ventures, Leitmotif, Matter Venture Partners, Mayfield, Premji Invest, Prelude Ventures, Temasek, and Xora, as well as technology investor John Doerr.
Rhoda AI is led by co-founder and chief executive Jagdeep Singh. The leadership team also includes chief science officer Eric Ryan Chan, a researcher in computer vision and generative modeling, and Gordon Wetzstein, a Stanford University professor who leads the Computational Imaging Lab. The company’s engineering team includes researchers working in generative artificial intelligence, computer vision, and robotics.
